Speaker
Description
The ePIC experiment at the upcoming Electron-Ion Collider (EIC) is advancing toward compute-detector integration with seamless data processing from detector readout to analysis. This paradigm shift in data processing is driven by streaming readout and AI technologies.
Streaming readout captures every collision signal, including background events, ensuring no information is lost. This holistic approach enables event selection by utilizing all available detector data, eliminating trigger bias, and providing accurate uncertainty estimation. Moreover, prompt background measurements through streaming readout play a critical role in reducing background noise and associated systematic uncertainties, pushing the boundaries of experimental precision.
AI plays a significant role in optimizing the data processing pipeline. Autonomous alignment, calibration, and validation help accelerate data turnaround. Beyond this, AI has the potential to enhance detector systems at ePIC through autonomous experimentation and control. For instance, a smart detector system could dynamically adjust thresholds based on background rates, ensuring optimal performance in near real time.
This talk will highlight streaming readout and AI applications at ePIC, showcasing their potential to redefine experimental capabilities and precision.
